Universal Access To All Knowledge
Home Donate | Store | Blog | FAQ | Jobs | Volunteer Positions | Contact | Bios | Forums | Projects | Terms, Privacy, & Copyright
Search: Advanced Search
Anonymous User (login or join us)
Upload

Reply to this post | Go Back
View Post [edit]

Poster: hseamons Date: Apr 28, 2012 9:50am
Forum: GratefulDead Subject: For cosmiccharlie - Post retirement matrix sample...

Greetings! One of the major parts of making matrixes is that there have been a lot of aborted matrixes over the years, due to the sync or mix not working out. This has been the case recently with 9/1/79 at Holleder Stadium - Rochester, NY. While syncing the first song, "Mississippi Half-Step Uptown Toodeloo," went fairly well (although being more difficult than average) things got really squirrely during "Franklin's Tower." While I could have cut out the AUD for the couple minutes of intense speed shifts that were occurring and try to resume the matrix later in the song, I opted to abort the matrix, seeing that the tape was not being very conducive to the process overall and that other issues would inevitably arise, it appears. I am blaming this problem on the multiple generation SBD source. Hopefully, a better quality transfer of the SBD will appear one day, so that I can resume work on what would be a very fine sounding outdoor matrix! While I hate to dispose of work that has already been done, here is a link to the entire "Half-Step" track for download (with part of "Franklin's Tower" in it), which is a really good version that is definitely worth the download:

Half-Step MTX, 9.1.79

Please make sure you click the "Click here to start download from sendspace" link, as there are other "fake" download buttons that will only confuse you. Thank you, and happy listening!

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 28, 2012 1:54pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

My suggestion is to manually pitch-correct the worst segments. I did this myself by marking in/out points for each source, and then measuring their differences exactly. I used the "difference data" to define the speed correction amount. Then I used Resample Mode to correct the speed and the pitch together. The s/w does the math for you. Repeat this process as necessary. In my case, the tapes I matrixed were way out of sync. Many longer songs on the SBD tapes required me to pitch-correct them in three-minute segments, until I reached the end of the track. It was a lot of extra work. However, if each source is of "listenable" quality, you can definitely sync them using my method. Want me to sync them for you?

Hunter, this is the s/w module ("app") that I used to pitch-correct the SBD to sync to my June 9, 1973 AUD tapes.
dead-head_Monte-samplitude-pitch-speed.jpg

in-depth discussion about Pitch Correction & Speed Correction

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 28, 2012 2:15pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

I appreciate the info and the offer, Monte, but let me think about it some. I'll let you know if I'm game.

In any case, isn't the idea to manually pitch correct just about every segment, not just the worst ones? I find this needs to be done in as short as 20 second intervals, but usually longer than that....Of course, this may have to do with how much control one has over the speed itself. I can adjust speed to the nearest hundredth of a percent in Final Cut Pro, which has seemed to be flexible enough, especially when working with good sources.

The problem, in this 9/1/79 case, seems to be that there is so much drastic speed change happening at a constant rate that it's hard to feel like things are stable, and it's going to degrade the sound in some form, despite efforts to correct it with our snappy editing programs. I guess if there's a will there's way, but the the weirdness in Franklin's Tower was beyond my capabilities....it's just not a normal moving tape speed - as it damn well should be!

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 28, 2012 3:04pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Often, two tape decks are off-speed in a linear way. Somebody like Joe B. Jones will do a pitch analysis for this case. The pitch-correction fix in this case is simple. The whole tape is pitch-corrected.

Other speed problems between tape decks involve "momentary" transport trouble. These are non-linear. Sometimes they come and go. Sometimes they get worse, and then better, and then worse again. Problems like these sometimes happen: battery power bad; shore power bad; power supply problem inside tape deck; capstan belt slipping; pinch roller is bad; mechanical brake is dragging; motor brushes bad; motor drive electronics went bad; tape packing on reels is bad; cassette housing bad; etc.

This post was modified by dead-head_Monte on 2012-04-28 22:04:42

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 28, 2012 10:00pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

So, are you saying you can manage to correct the non-linear speed problems that are occurring, which is what I think I'm experiencing? I'm not sure if it's the AUD or the SBD, to be honest, but I can not rule out either. I've experienced some wacked out SBD sources, for sure, because I've tried different SBD transfers at times that have eliminated certain problems.

Actually, I think it is the SBD because I could tell things were getting slower and slower in a weird way, and the AUD was the "pitch corrected" source - so it has to be the SBD

This post was modified by hseamons on 2012-04-29 05:00:35

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 29, 2012 6:21am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Yes, I am saying this, for example:
a) Point out two "listenable" (good quality) sources to me that you want me to match and sync together. One of these sources will need to get pitch-corrected to the other. I will correct the non-linear speed-error portion(s), and match both sources until they sync together.

b) How do I do this? I select the first segment of a track that I cannot easily sync with another source. In this example, let's say I'm correcting segments that are three minutes long, and I'm working down the track, starting from the first note of the tune Sugar Magnolia.

c) For both sources, I locate and mark the exact in-point for the first note. The exact same point is marked on both sources. Then I jump down the track by about three minutes. I search the track's content for a note or sound that is familiar to me. I scrub the tape until I mark the exact out-point for this familiar note. I locate the exact same out-point on the other source, and I mark it.

d) Using my very accurate in/out points, I measure the exact duration of this "three-minute segment" for both sources. What are their durations, exactly? What is the speed difference measured between them, exactly? This is the "difference data" I need. There is NO Guessing work involved with my method.

e) I convert the speed "difference data" into a ratio. This number becomes the Time Factor number that I dial into the Resampling Module. I set this module for Resample Mode. Then I apply the pitch and speed correction to the one source.

f) Using Samplitude Pro VIP mastering suite, I independently position, sync, and mix the two sources until they are correctly matrixed.

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 8:11am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Thanks much for the info, Monte. I was thinking you might have had a method for correcting non-linear problems without having to compare sources. Doesn't Plangent Processes do that? I have encountered many non-linear issues in my time making matrixes, which I have been able to work through, but this instance was particularly difficult and annoying, and I was not digging what was happening with the SBD tape during certain parts.

In any case - and not to sound like an arrogant bastard here or anything - but after reading your syncing methods carefully, I honestly think they are not refined enough. Just my 2 cents.

Thanks again for your input and willingness to help.

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 29, 2012 9:20am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

What's required is an Audio Engineer applying Character Animation techniques to the tape reels that are bad. The best way to approach this is to use a Keyframe Editor. My s/w doesn't have one. You also need non-linear speed controls for doing ease-ins and ease-outs. My s/w doesn't have these. It does Linear correction ONLY. I have discussed all these factors in my pitch-correction threads that I linked for you.

In the meantime, I have already proven that my methods and techniques for doing these "manual tape speed over-rides" are workable for people like yourself. This is a labor-intensive work-of-love that is a "cheap alternative" to spending huge sums of money on expensive equipment. (My semi-professional conclusion is just that - it's semi-professional.) By exactly comparing the sources, I can exactly measure the differences - like a car mechanic (or a character animator). I agree with you that matching three-minute segments is a "low resolution" approach to matching sources "perfectly". I was intimating that I am less-workaholic than you. I used the "bare-minimum method" so-to-speak. But, when I matrixed June 9, 1973 this way, I had great quality AUD and SBD sources to work from. (The opposite of this is... Garbage In = Garbage Out.)

We are agreeing in principal. I am giving away for free, information on how to do this, for anyone out there that might be inclined to do this work. The existing "public" GD library of vintage tape sources has been scoured and mined for its matrixable sources. It looks like the matrix-buzzards picked through all the bones of this carcass. I have a special matrix-buzzard "speed-fixing" tool to offer people who need a matrix-fix, carpal-tunnel syndrome not withstanding.

http://archive.org/details/gd1973-06-09.117476.mtx.barry.flac16

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 9:49am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

I have very often noticed in my audio edits, based on the edit lines I create, certain patterns that reflect "ease ins" and "ease outs," so I think I've managed to accomplish something in terms of handling non-linear anomalies - it's very similar to keyframing, but it looks more like a "connect the dots" line than an automated curved line that certain keyframes would create. In any case, since we view images nowadays that are essentially composed of squares (pixels), yet look incredibly smooth, I think the same can be accomplished with a "high-resolution" editing style, using linear patterns at a very short rate, that the human ear essentially cannot detect when blended with a pitch-corrected recording. That's why I like to call them Trix.

Anyways, you have inspired me to give 9/1/79 another shot - at a higher resolution........

peace

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 10:11am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Actually, now that I think about it more, my "connect the dots" line is really more of a curved line by virtue of how I respond to the rate of speed change that is happening between each dot, or "keyframe." In other words, the lines between the dots are actually more curved than straight because speed-change is ultimately happening in analogue recordings - which is why using analogue is so essential to this process, and why I'm so glad these things were put onto tape in the first place.

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 29, 2012 2:17pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

By exactly measuring the sources, the exact speed differences become crystal clear to you. You are the speed-control animator. As you go down the tape, you will be witnessing what the trending is, by measuring all these speed differences. For example: a) fast, faster, and faster; b) slow, slower; and slower; c) faster, slower, then faster; d) slower, faster, then slower; e) cyclical speed-changing patterns; f) "once-around" errors; g) start-of-tape errors, then middle-of-tape is good, then end-of-tape errors.

For Samplitude Pro s/w, the pitch correction module needs to be integrated into the VIP mixing, mastering, and rendering module. This enables you to preview your pitch and speed shifts while you mix all your sources at the same time. Then you need a robust Keyframe Editor to create animation "fairings" that match the speed-errors exactly, including automation tools like these:

Easy Ease — Smooths both the keyframe’s incoming and outgoing interpolation
Easy Ease In — Smooths the keyframe’s incoming interpolation
Easy Ease Out — Smooths the keyframe’s outgoing interpolation

09fig80.jpg

Adding slight deceleration to soften or ease the speed into and out of keyframes is a commonly used technique. Adobe is applying Keyframe Assistants "apps" in After Effects. These are provided to automate the task of keyframe smoothing.

Animating on the Fly in Motion 4
S0710_AnimFly.png

A powerful feature that Motion offers is the ability to create animations while your project is playing back. This is similar to how audio engineers adjust sliders for each of their different audio channels while listening to the mix. The next time it is played back, all of those changes are incorporated.

Because so many of Motion’s effects are displayed in real time, you can perform a sort of “visual mix” and modify the various parameters of your effects while the project is playing back.

(Imagine seeing all the tape segments strung together in a Keyframe Editor. You're in the middle of a pitch-correction project. For the current song you are working on, now you can know whether each part is behaving as it should.)

With keyframing on the fly, you can make adjustments to any slider or parameter in the program while your project is playing. You can also interactively manipulate objects in the Canvas. Then, if you don’t like what you did, rewind and do it again. Each time you alter a parameter, you replace any keyframes previously assigned.

One of the most common uses for animating on the fly is to create handmade animation paths. Although you can choose from a variety of curve types, it is very difficult to get the semi-random movement except by sketching with your mouse or pen.

This is an animation (of motor-speeds) issue. Aircraft accident reconstruction animations address this. Forensic re-enactment animations used in courtrooms address this. Car accident reconstruction animations address this. Architectural Designer CAD s/w addresses this. They all use the same technology. Technicians, engineers, and scientists measure stuff with precision, and then they create computer animations that visualize exactly what they measured. This process is called Scientific Visualization. (The purpose of scientific visualization is to graphically illustrate scientific data to enable scientists to understand, illustrate, and glean insight from their data.) Doing a Matrix this way is like performing a Scientific Audio Rendering.



This post was modified by dead-head_Monte on 2012-04-29 21:17:59

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 10:53am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Got it. My only concern, when it comes to the subjective art of sound, is that super-scientific methods might overshadow an alternative, more enjoyable acoustical energy that can be achieved by way of the mix in tandem with the ACTUAL, MINUSCULE distance between the "present moment" of the SBD and the "present moment" of AUD, which I think is a balancing act from show to show, song to song. They should not be right on top of each other, and they should rub each other the right way, like in good sex. What I'm saying is that there should be room for "play" rather than some theory of sonic "perfection," and that mathematical calculations might create limitations - certainly they are not always the right ones for the task at hand. I can understand the need to apply these techniques to car accidents, but I would prefer much more caution with matrixes, which to me are an art form more so than a science. I don't like using targeting computers when going in for the kill - that's almost like cheating.

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 29, 2012 11:45am
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

The art here is purely scientific. Take primary colors, for example. The combination of any two exact primary colors creates an exact secondary color, every time. But, it takes a long time to mix colors and create a good painting. Translated: the combination of any AUD and SBD, pitched and synched exactly together - taped at the same show, creates an exact matrix, every time. But, it takes a long time to create and mix a good matrix.

The sound engineer knows that when mixing two exact sounds (frequencies) together, the result is always four exact frequencies. You will always have the two original frequencies, the sum frequency of the two, and the difference frequency between the two. Every time. For the AUD, you will ALSO have combinations of reflected and delayed sounds. On the SBD side, some instruments may be delayed differently than others, especially when using effects boxes. The P.A. system may also have sub-system arrays that are delayed differently from each other. Alembic did this with GD's sound system during the wall of sound era. AUD and SBD sources will mix differently into each matrix attempt.

Now it's an 'art' to get the matrix sounding the best.

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 2:03pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Are you suggesting that the SBD and AUD need to match exactly, as in they need to be synced right on top of each other? I think the AUD should generally trail behind the SBD ever so slightly. This creates a more "complementary" sound, from my experience. This distancing, which is what I am getting at, is a creative, subjective choice, not a scientific one. What number do you choose for this distance? "3.41"? "3.75"? It's more subjective than one would like to think. On top of that, I think automated "ease ins" and "ease outs" will not match the other tape 100% perfectly - but it can come VERY close. There's always room for minuscule error when applying "samples" to something that is by nature "organic," even with our best instruments. There's always a bigger microscope. The beauty is, however, we can have a "just exactly perfect" sounding blend without it matching perfectly. It's a matter of: how close to the painting do you really need to get? You ultimately have to stand back from it to focus on the experience itself, not the "perfect" or "semi-perfect" brush strokes, which are more or less hidden from the human ear anyway, if done well.

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 2:54pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

On second thought, I suppose you could just sync it all on top of each other using the techniques you have described, and then just push the AUD back however far is desired, as one big piece. (But something tells me that might not work across the board.) So, yeah, the more technology to figure this stuff out the better. I guess we are in total agreement at this point. Great discussion!

Reply to this post
Reply [edit]

Poster: dead-head_Monte Date: Apr 29, 2012 2:55pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

No, I'm not saying whether to use synch-offsets "for effect" or not. Synching two tapes will never be perfect for the whole show in a matrix project. But you can get close, like shooting fish in a barrel.

Matrixing and synching messed up tapes is extremely difficult. What if both sources are messed up? It is very time-consuming, including carpal-tunnel syndrome. I'm simply saying, you can get your ducks in a row, and have a life at the same time. Don't kill yourself, and don't humiliate yourself. Consider the science involved. If you do end up quitting this project, at least you'll be an expert at it.

What do you think of the "Bertha remaster" sound, like this Veneta show that Jay Ashley did? I like it! It's not a matrix, or synch-offset, but it's a DAW (digital audio workstation) production. Similar to the matrix controversy, some people like the Bertha sound, and some dislike it.

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 29, 2012 3:07pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Got it, makes sense. Use the best tools possible! If they really make it work better, then that's the way to go. Thanks for sharing all the info.

The Bertha's are tricky, because it's totally subjective. Some will like them, others won't. That's just how I feel about the Bertha's: I like some, others I don't so much.

Sometimes I like hearing the "classic" GD recording sound over a more "produced" sound - but man, sometimes that produced, "enhanced" sound just does it for me.

Reply to this post
Reply [edit]

Poster: Skobud Date: Apr 28, 2012 1:26pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Hi Hunter, you nut.

How have you been? I read your post about stopping and I dont think you should. You contribute, and that is always a good thing. Anyhow, I cant get this link to work. Would you consider using rapidshare? Its easier and better than sendspace imo. You got me interested in this thing and I cant grab now. Says free capacity reached or some shit. Thanks.

Reply to this post
Reply [edit]

Poster: hseamons Date: Apr 28, 2012 2:11pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Thanks Skobud. I looked into rapidshare, and it looks like you can only share with other people who have rapidshare, as it does not create a direct download link - at least in the free mode. But maybe I'm missing something?

The only reason why I started this matrix is because my wife was out of town for a few days, and I knew I would not end up in any heated arguments as the result of blowing up the apartment with Grateful Dead music from the Brent era. That's how difficult this process gets for me - and why I still consider myself retired as of 4/1/12.

Reply to this post
Reply [edit]

Poster: cosmicharIie Date: Apr 28, 2012 2:29pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

for gods sake Hunter, my name is spelled with ONE c! and a big~eye, do Tell! btw, thanks, nice half-step, too bad about the source tho now to continue the darkstar to blackhole tour (no, not some hollow), i booted rmon into one...Waggyner most have had the jiiitters

later cc

Reply to this post
Reply [edit]

Poster: William Tell Date: Apr 28, 2012 4:21pm
Forum: GratefulDead Subject: Re: For cosmiccharlie - Post retirement matrix sample...

Wait a minute...oh, never mind...damn, this is confusing.